Sonora
🎙️
Streaming dictation
Text appears as you speak (with Gemini Live).
Open source · 100% built with AI
Speak. The text writes itself.
Sonora is a Spotlight-style floating bar, always a shortcut away. Press, dictate, and the text appears live — copied to the clipboard or typed straight at your cursor.
Everything you need to dictate fast
A minimalist tool, designed to stay out of your way.
⌨️
Type at cursor
The result is typed where you are, or copied to the clipboard — your choice.
⌘
Global shortcut
Start and stop dictation without ever leaving your current app.
🔌
Pluggable engines
Gemini, Mistral (Voxtral), OpenAI, Groq, any compatible endpoint, or local Whisper.
✦
Hesitation cleanup
An optional pass removes "uh", false starts and repetitions.
✍️
Reformulation prompts
Transform a dictation through an LLM: formal tone, terminal command, professional rewrite…
🕘
History
Find and re-copy your previous dictations in one click.
🔐
Keys in the keychain
Your API keys are stored in the OS keyring, never in plaintext on the interface.
🌗
Theme & waveform
Light, dark or system, with a sound wave driven by the real audio level.
An engine for every need
From the fastest cloud to 100% offline. You stay in control of your data.
| Provider | Transcription | Reformulation | Key required | Offline |
|---|---|---|---|---|
| Gemini Live | streaming word-by-word | ✓ | yes | — |
| Mistral (Voxtral) | per segment | ✓ | yes | — |
| OpenAI Whisper | per segment | ✓ | yes | — |
| Groq Whisper | per segment fast | ✓ | yes | — |
| OpenAI-compatible | per segment | ✓ | host-dependent | possible |
| Local Whisper (ggml) | per segment | — | no | ✓ |
Three steps, that's all
From install to your first dictation in under two minutes.
Install & choose an engine
Download Sonora, open the Settings (⚙ icon on the bar), pick a provider and paste your API key (or point to a local Whisper model).
Press the shortcut
The floating bar appears and dictation starts. A sound wave confirms Sonora hears you.
Text lands at your cursor
Release: the text is typed where you were working, or copied to the clipboard. Reformulate it if needed with your prompts.
Sonora in action
A demo is worth a thousand words: the dictation writes itself live, then lands at the cursor.
Frequently asked questions
Everything you need to know before your first dictation.
Is Sonora free?
Yes — Sonora is 100% free and open-source under the MIT license. No account, no subscription. Bring your own API key for cloud engines, or run everything locally for free with Whisper.
Does Sonora work offline?
Yes. With the built-in local Whisper engine (ggml), transcription runs entirely on your machine — no API key, no internet, and nothing leaves your computer.
Which platforms does Sonora support?
macOS (Apple Silicon & Intel), Windows and Linux. Sonora ships as a lightweight native binary built with Tauri.
Is Sonora an open-source alternative to Wispr Flow, superwhisper or Dragon?
Yes — Sonora is a free, cross-platform, privacy-first alternative to voice-dictation tools like Wispr Flow, superwhisper and Dragon NaturallySpeaking. You choose your engine (Gemini, Mistral, OpenAI, Groq or any OpenAI-compatible endpoint) or run fully local Whisper.
Are my API keys and dictations private?
Your API keys live in the OS keychain, never in plaintext. Audio is sent only to the provider you pick — or nowhere at all with local Whisper.
Can Sonora type directly where I'm working?
Yes — Sonora types the text straight at your cursor, or copies it to the clipboard. On Linux it uses wtype (Wayland) or xdotool (X11); on macOS it needs the Accessibility permission.
Ready to dictate?
Open source, free, MIT-licensed. Your keys and your data stay with you.